Hi,
It might be worth checking if the input images contain NaN values, as these will cause FSL programs to fail. One exception is fslmaths, that can be used to detect and remove NaN values using the -nanm and -nan options respectively. If there are no NaNs or this does not help, let us know.
All the best,
Mark
On 16 Sep 2013, at 12:31, "Hellyer, Peter John" <[log in to unmask]> wrote:
> Hi All,
>
> I'm trying to run fslvbm on a collection of healthy controls, but there appears to be something strange happening at the second stage fslvbm2.
>
> when I look at template_4D_GM there are a lot of subjects (31) that are completely blank, and when I look at the associated first segmentation images, the images are all ones.
>
> I had a look at the the error output of fslvbm2a, which reveals the error 'MeaNsK variance nan' fslvbm2b produces the warning 'WARNING::in calculating COG, total = 0.0'
>
> which suggests some kind of fault with the input data, but I can't see anything obviously wrong. step 1 completes okay, and the betted images look fine.
>
> Any suggestions as to how to proceed would be welcome
>
> Best Wishes,
>
> Pete
|